Colt's Blog

Luke Broome, Colt’s CTO, recently said “Investing into network programmability is an important step towards delivering integrated network and compute services and helps keep Colt at the forefront of innovation. This evolving model brings the network closer to the applications and will enable our customers to flex and consume their managed IT and network services in new ways.”

After keeping an eye on the developments in the area of software-defined networking (SDN) in recent years, working with partners on defining use-cases and executing proof-of-concepts we decided to go “public” and join the Open Networking Foundation back in March this year. We have a long history of driving innovation in a number of areas like Carrier Ethernet, managed services and modular data centres and plan to push network programmability with the same energy!

But let’s be clear, this will be a journey: an exciting one for sure, but potentially a long one. Most disruptive innovations are built on previous innovations and aren’t completely new. If you’ve been in the industry for long enough, you probably got the feeling of “I’m sure I’ve seen this somewhere before…”. I know I did in the early days. So what’s different this time, what makes SDNs so revolutionary? In my opinion it starts with the fact that it caused people to think differently about how we build networks, bind them with applications and manage them in-life. And that’s exactly what we need: not just a protocol (e.g. Openflow) or a toolbox (e.g. APIs) but a new end-to-end integrated architecture to deliver, operate and consume combined network and managed IT services. And why do we want to do this? Reducing the cost of the overall solution might be one of the outcomes, but this may eventually be overshadowed by the benefits that this journey brings to enhance the customer experience. And that is a very exciting prospect.

In the coming months you’ll hear more from us on how we build on this new paradigm to deliver network automation and network virtualisation, how our data centre fabric will evolve and our plans for wide area networks. Watch this space!

Since 2010’s infamous “flash crash”, High Frequency Trading (HFT) has been under the continuous gaze of the financial services regulatory authorities. HFT's ability to enable large numbers of high speed trades made it the prime suspect when the Dow Jones Industrial Average plunged 1,000 points in a matter of minutes, before quickly recovering. According to some market watchers, HFT increases volatility and damages the integrity of the markets. From firms being fined millions of dollars for computer errors disrupting the futures market, to commodity traders getting stung for market manipulation and alleged mini flash crashes like the case of Thermo Fisher Scientific and Pall Corporation, in which shares dropped by more than $1 before regaining most of that value in minutes, the practice has come under increasing scrutiny.

A recent report by the Futures Industry Association (FIA), however, shows that instead of contributing to the Dow Jones 1,000 points fall, high speed trading actually absorbed the initial sell orders, with the majority of algo traders continuing to provide market liquidity throughout the drop. Most importantly the study confirms that speed is an essential risk management tool for market participants, specifically electronic market makers. It states that speed enables many organisations to post direct quotes at competitive prices on exchanges, safe in the knowledge that the quote can be updated very quickly if the market moves.

The report’s findings should hardly come as a surprise. In the complex world of HFT, speed, as we have highlighted in previous posts, is of paramount importance in the race for market liquidity. With the majority of firms currently adapting highly sophisticated algorithms to search the stock market for tiny price differentials, trading is no longer about who spots the profit opportunity, but about who gets there first. This is why proprietary trading shops are continuing to invest in upgrading their IT infrastructure and to pay for co-location, putting their servers as close as they can to the local exchange’s matching engines. It is also why these firms invest significantly to maintain the fastest fibre-optic connectivity in a bid to reduce latency, enabling their traders to cut a few milliseconds off order execution times.

This practise is not restricted to Europe alone. The emergence of new markets catering to cross venue arbitrage across the world means that trading firms will have to adapt. If firms are to take advantage of arbitrage - price differences between two or more markets- it is critical for them to have their trading engines physically present in emerging market cities, so that they can exploit price fluctuations. An optimally-located trading facility is an advantage over rivals, who will face greater delays in obtaining information from either one exchange or the other.

FIA’s conclusion that high speed trading has a positive impact on markets should provide a prompt for the industry as a whole to focus on HFT not as something for the few, but as a practice for any trader looking to use technology looking to get faster access to market liquidity. Hopefully, the debate can finally move on from the effects of HFT, to the potential of new technology to progressively raise trading speed and agility.

Recently at the Tier1 Research Hosting and Cloud Transformation Summit in London, I sat on a panel discussing the EU legal and compliance landscape with the CTO EMEA of Microsoft, Stephen McGibbon, Digital Realty Trust VP Adam Levine, Hogan Lovells’ Partner, Conor Ward and Tier1’s Daniel Beazer.

Within the last few weeks, Hogan Lovells has published a further paper on this issue, specifically regarding the Patriot Act and its equivalents in Europe entitled “A Global Reality: Governmental Access to Data in the Cloud”. The White Paper is a very good read. It challenges the frequently-expressed assumption that the United States is alone in permitting governmental access to data for law enforcement or national security reasons. It examines the laws of ten countries, including many in Europe, with respect to governmental authorities’ ability to access data stored in or transmitted in those countries, and documents the similarities and differences among the various legal regimes.

The conclusion is that US requirements are not that different from the rest of the world. As the report says “Governmental access to data stored in the Cloud – including cross-border access – exists in every jurisdiction”.

On the other side of the pond, critics of the EU's data protection regime like to characterise it as a form of protectionism designed to make entry into the European markets difficult for outsiders. But concerns about this and the Patriot Act were succinctly put into context by Conor Ward at the Tier1 panel session. According to Ward, it has more in common with the practise of hiding behind Health and Safety as a reason for inactivity - a blanket excuse not to perform a difficult or inconvenient task.

He argued that scaremongering is rife in the market, pointing out that cloud computing is not unique or new and neither are privacy issues. Likewise, offshore and outsourcing are decades old. He went on to state there is nothing new, unreasonable or onerous about the regimes currently in place; outsourcing has existed for decades and therefore the legal framework that surrounds moving data is relatively mature.

During the recent VMware Forum event in Manchester, I was lucky enough to be invited to contribute a keynote speech on Hybrid Cloud and more specifically, how it relates to the channel. You can view the presentation that I gave below, but I would like to explore here some of the themes I touch on within it. The first point I made to my audience was that from a market perspective, the demand for Infrastructure-as-a-Service has seen dramatic growth in the last three years with a 50% growth rate. The IaaS market itself has seen demand for Compute grow by 69% and Storage grow by 60%. The IT requirement underpinning this is need for elasticity in on-demand services. Capacity and scalability need the utmost flexibility.

Resellers are beginning to realise this and change the mix of their sales portfolio, focussing more on services. The channel is now realising that cloud is a customer intensive consultative process. The customer wants to be able to “flick a switch” to turn services on and off, but this is often tricky to achieve and needs the help is an engaged and smart reseller. I take a point from Nicholas Carr’s book The Big Switch that “Operating your own IT assets is becoming akin to running your own power station”. Computing capacity is becoming another utility and the decision to run internal personnel and technology will make less and less business sense as time progresses.

The main advantage from the hybrid cloud position is therefore one of saving costs, matched with efficient time for delivery and implementation while not compromising quality. The quality of the service is vital and at Colt it’s what I believe we do best. This approach isn’t about cutting-corners or matching the expectations around publically managed cloud services. It’s about delivering cloud service provision with flexibility, consistency and expert management. Take a look at the video below (or click through it in the presentation) to our case study interview with Michael Meaney of the Smart Gaming Group who’s kindly provided us with an excellent overview of what Colt hybrid cloud did for him. With solutions like this available, it’s time for IT directors and resellers alike to question where they are in the IT provision space, and to consider where is their management time best invested.

At Colt we’re improving the service offering all the time and finding new applications. Our new Disaster Recovery system, powered by vCloud technology, addresses a fundamental business challenge of keeping your stored data, processes and infrastructure safe should the unthinkable happen. Few companies can afford the dedicated, hi-quality standby facilities which their business users think that they ought to have. With vCloud from Colt, hybrid architectures dramatically improve the quality of DR provision available to customers. The possibilities are only just beginning to be realised and we intend to bring more services to the channel in this dynamic area.

Demand for data storage is growing at a tremendous rate in European markets such as France, where Colt recently made an additional 1,000 m2 of modular data centre space available at its flagship site South of Paris.

Exponential growth of data means that the modern business organisation must have cutting edge access, speed and security simply to stay competitive in a global market. Fortunately, France is well placed to meet this demand with a data centre sector primed for expansion thanks to its secure electricity supply and the high calibre of its engineering skills base.

French legislation, in particular, provides for the security of data and data processing while, at the same time, meeting industry needs for a secure legal framework. Indeed, according to the Invest in France Agency (IFA), estimates indicate that investments in data centres in France will reach €1 billion a year, having supported more than 26 data-centre projects since 2006.

The significant challenge facing the sector, however, is the need for easily scalable solutions to meet businesses’ growing data storage demands. The Cloud and new web-oriented, virtualised applications, indeed the very nature of applications themselves is constantly changing, now more than ever.. Data centre capacity planning in this fast paced environment is difficult with a solution that can take years to build. This is where the traditional data centre is struggling to keep up.

Being able to effectively manage capacity and therefore capital expenditure and operational expenditure is a distinct competitive advantage. Commissioning a new data centre in the traditional way is no longer the answer for businesses operating in extremely competitive local and international markets.

Modular approach

By combining the modular concept with flexible design variations, Colt’s approach allows complete data centres with industry-leading efficiencies to be installed and fully commissioned within just sixteen weeks, thereby eliminating most of the uncertainty around capacity planning.

The modular approach provides customers with higher flexibility than ever before in power, size and infrastructure capacity previously only associated with bespoke builds. This is achieved using modern production line techniques to manufacture an extensive range of standard data centre variations including options for layout, floor size, power density, cooling technology and tiering.

By adopting a manufactured approach, the performance and reliability of the completed hall can be precisely predicted and optimised. The result is large-scale data centres that match or exceed the durability and quality standards of a traditional data centre, but are ready for work in a fraction of the time.

Improved power efficiency

The new data centre in Paris is a case in point with a proven industry leading design Power Usage Effectiveness (PUE) of 1.21, representing a significant improvement over many traditional data centres and could result in significant cost savings for customers. Some traditional data centres have a PUE as high as 2.50 and based on these levels, a 500m2 modular data hall centre with an IT load of 750kW and taking into consideration inflation at five percent per annum, could result in a cost saving of 3.6 million Euro, the equivalent of over 10 years in power bills in France.

Improved reliability and reduced downtime

Since data centres are the heart of today’s business processes, a single downtime event has the potential to impact significantly on the profitability of an organisation. But, it is not only the obvious initial cost that is a burden. There are far reaching repercussions: the cost of fixing the problem, the data recovery and, most importantly, the potential damage to the company’s reputation which could cost millions in future business.

The Colt technology at the core of our Paris data centre has been developed to offer exceptional standards of performance and reliability. Standardised, proven components, factory built and tested into designs that are themselves standardised can offer performance which is absolutely predictable.

This is supported by a sophisticated system architecture, which will keep the system running even in the event of the failure of individual elements. Protecting it all is state of the art power supply, fire protection and security systems. It all contributes to the reliability that any modern data focussed operation must have.

The business case for a modular data centre

The greatest benefits of utilising a modular data centre are actually commercial rather than technical. The costs of design and build, capital and time operations, power and downtime add up to improved total cost of ownership, offering early adopters clear competitive advantages. The modular approach provides several key advantages over conventional builds:

• It can provide a new data centre within 16 weeks
• It allows the use of standardised, highest-quality components
• It ensures that cutting edge technology can be used in both servers and ancillary systems
• It provides a proven design with predictable performance
• It provides the flexibility to grow easily with your demand for capacity